Search Images Maps Play YouTube News Gmail Drive More »
Sign in
Screen reader users: click this link for accessible mode. Accessible mode has the same essential features but works better with your reader.

Patents

  1. Advanced Patent Search
Publication numberUS8031775 B2
Publication typeGrant
Application numberUS 11/347,679
Publication dateOct 4, 2011
Filing dateFeb 3, 2006
Priority dateFeb 3, 2006
Also published asUS20070182861
Publication number11347679, 347679, US 8031775 B2, US 8031775B2, US-B2-8031775, US8031775 B2, US8031775B2
InventorsJiebo Luo, Christophe E. Papin, Deniz E. Schildkraut, Majid Rabbani
Original AssigneeEastman Kodak Company
Export CitationBiBTeX, EndNote, RefMan
External Links: USPTO, USPTO Assignment, Espacenet
Analyzing camera captured video for key frames
US 8031775 B2
Abstract
In order to analyze a digital video clip to determine candidate frames for subsequent key frame selection, a camera motion sensor is provided in the camera so that information is provided during image capture regarding camera motion including translation of the scene or camera, or scaling of the scene. The sensor includes an accelerometer or a lens motor sensor. A plurality of video segments is formed based on the global motion estimate and each segment is labeled in accordance with a predetermined series of camera motion classes. Thereafter, key frame candidates are extracted from the labeled segments and a confidence score is computed for each candidate by using rules corresponding to each camera motion class and a rule corresponding to object motion.
Images(7)
Previous page
Next page
Claims(22)
1. A method of analyzing a digital video clip captured by a camera to determine candidate frames for subsequent key frame selection, comprising:
a. providing a camera motion sensor, including an accelerometer or a lens motor sensor, in the camera so that camera motion information is provided during image capture regarding camera motion including translation of a scene or camera, or scaling of the scene;
b. forming a plurality of video segments based on the camera motion information and labeling each segment in accordance with a predetermined series of camera motion classes; and
c. extracting key frame candidates from the labeled segments and computing a confidence score for each candidate by using rules corresponding to each camera motion class and a rule corresponding to object motion.
2. The method of claim 1 wherein the predetermined camera motion classes include pan (left or right, and tilt up or down), zoom (in or out), fast pan or fixed.
3. The method of claim 2 wherein the rules include a pan rule, a zoom rule, a fast pan rule and a fixed rule.
4. The method of claim 3 wherein the pan rule includes extracting a plurality of frames to cover the space of environment while reducing the spatial overlap among the frames from a pan segment.
5. The method of claim 3 wherein the pan rule includes extracting a frame located at a point when the pan motion is slowed down.
6. The method of claim 3 wherein the zoom rule includes extracting a candidate frame at an endpoint of the zoom-in or zoom-out segment.
7. The method of claim 3 wherein the fast pan rule includes extracting candidate frames from a fast pan segment.
8. The method of claim 3 wherein the fixed rule includes extracting a candidate frame located at a midpoint of a fixed segment.
9. The method of claim 1 wherein the object motion rule includes extracting a candidate frame for a fixed segment with a confidence score related to the segment length, extracting a candidate frame for a pan segment with a confidence score related to a translation amount during the pan, and does not extract candidate frames for object motion based frames for fast pan and zoom segments.
10. A method of analyzing a digital video clip captured by a camera to determine key frames, comprising:
a. providing a camera motion sensor, including an accelerometer or a lens motor sensor, in the camera so that camera motion information is provided during image capture regarding camera motion including translation of a scene or camera, or scaling of the scene;
b. forming a plurality of video segments based on the camera motion information and labeling each segment in accordance with a predetermined series of camera motion classes; and
c. extracting key frame candidates from the labeled segments and computing a confidence score for each candidate by using rules corresponding to each camera motion class and a rule corresponding to object motion;
d. selecting key frames from the candidate frames based on the confidence score of each candidate.
11. The method of claim 10 further including ranking the selected key frames in accordance with the confidence score.
12. The method of claim 11 wherein step d includes employing the ranking and a user specified number to select the key frames.
13. The method of claim 12 wherein employing the ranking and a user specified number to select the key frames includes selecting at least one key frame from each segment if there are confidence scores above a pre-determined threshold.
14. The method of claim 13 wherein employing the ranking and a user specified number to select the key frames includes selecting key frames from the remaining candidates with the highest confidence values to fill the user specified number of key frames.
15. The method of claim 10 wherein the predetermined camera motion classes include pan (left or right and also tilt up or down), zoom (in or out), fast pan or fixed.
16. The method of claim 15 wherein the rules include a pan rule, a zoom rule, a fast pan rule and a fixed rule.
17. The method of claim 16 wherein the pan rule includes extracting a plurality of frames to cover the space of environment while reducing the spatial overlap among the frames from a pan segment.
18. The method of claim 16 wherein the pan rule includes extracting a frame located at a point when the pan motion is slowed down.
19. The method of claim 16 wherein the zoom rule includes extracting a candidate frame at an endpoint of the zoom-in or zoom-out segment.
20. The method of claim 16 wherein the fast pan rule includes extracting candidate frames from a fast pan segment.
21. The method of claim 16 wherein the fixed rule includes extracting a candidate frame located at a midpoint of a fixed segment.
22. The method of claim 10 wherein the object motion rule includes extracting a candidate frame for a fixed segment with a confidence score related to the segment length, extracting a candidate frame for a pan segment with a confidence score related to a translation amount during the pan, and does not extract candidate frames for object motion based frames for fast pan and zoom segments.
Description
CROSS REFERENCE TO RELATED APPLICATIONS

The present application is related to U.S. patent application Ser. No. 11/346,708 filed Feb. 3, 2006 entitled “Extracting Key Frame Candidates From Video Clip” by Jiebo Luo, Christophe Papin which is now U.S. Pat. No. 7,889,794 B1.

FIELD OF THE INVENTION

The invention relates generally to the field of digital image processing and, more particularly, to a method usable in extracting a key frame from a video clip.

BACKGROUND OF THE INVENTION

Extracting key frames (KF) from video is of great interest in many application areas. Main usage scenarios include printing from video (select or suggest the best frames to be printed), video summary (e.g. watch a wedding movie in seconds), video compression (optimize key frames quality when encoding), video indexing, video retrieval, and video organization. In general, key frames should show good quality and high semantic interest. However, what exactly is a key frame sometimes depends on the application. The level of requirement can also be different. For printing still pictures from video, one needs to put a strong focus on image quality. For rapid browsing one will need to increase the representativeness in semantics. Key frame extraction can be a feature offered in a camera (including a digital camera, camcorder, and camera phone), in desktop image/video editing/management software, and with online image/video service.

Key frame extraction is not a new problem. However, prior art has been focused on sports or news video with constrained structures. Such video conforms to well-defined common structures and characteristics. For instance, in field sports (including soccer, football, baseball, rugby, and cricket), there are two opposing teams and referees in distinct colorful uniforms, an enclosed playing area on grass or artificial turf, field lines and goals, commentator voice and spectator cheering, and finally, on-screen graphics (scoreboard). There are often a small number of canonic “views”: field view, zoom-in, and close-up. Other types of sports, such as racquet sports, basketball, as well as news videos, share a different set of structured characteristics. More importantly, there is unambiguous ground truth as to which are the key frames within the given context. In contrast, even the themed consumer videos (e.g., wedding, birthday party) do not have the same level of common structures and characteristics, and the key frame selection is open to a high level of subjectivity because of observer association, sentimental values, and other factors.

In addition, image quality (contrast, exposure, camera shake) is rarely a concern for sports and news video because of superior imaging equipment and well-controlled imaging conditions. Example systems for extracting key frames from sports and news videos include Avrithis, Y. S., Doulamis, A. D., Doulamis, N. D., and Kollias, S. D., “A Stochastic Framework for Optimal Key Frame Extraction from MPEG Video Databases,” Computer Vision and Image Understanding, 75(½), 1999, pp. 3-24; Liu, T., Zhang, H. J., and Qi, F., “A novel video key-frame-extraction algorithm based on perceived motion energy model,” IEEE Trans. Cir. Sys. Video Techno, 13(10), 2003, pp. 1006-1013; Y. Rui, A. Gupta, and A. Acero, “Automatically extracting highlights for TV Baseball programs,” ACM Multimedia 2000, pp 105-115; B. Li and M. I. Sezan, “Event Detection and Summarization in Sports Video,” IEEE Workshop on Content-based Access of Image and Video Libraries (CBAIVL), 2001, pp. 132-140.

Short movie clips captured by a digital camera with video capabilities (a recent product feature) are different. The variety in occasions and situations for consumer videos is unconstrained. Contrary to professional videos, there are no special effects, no tightly pre-defined structure, no professional editing, and a video clip represents only one shot. In that sense, video summary from a short clip is potentially easier than for those recorded by a camcorder because one does not need to perform video shot segmentation. Camera shake is often present and exposure is often problematic compared to professional videos. Above all, the biggest challenge with consumer video is its unconstrained content and lack of structure. Tong Zhang, in US patent application publication US 2005/0228849, “intelligent key-frame exaction from a video”, described a method for intelligent key frame extraction for consumer video printing based on a collage of features including accumulative color histogram, color layout differences, camera motion estimation, moving object tracking, face detection and audio event detection. Specifically, Zhang disclosed a method for extracting a set of key-frames from a video, comprising the steps of: selecting a set of candidate key-frames from among a series of video frames in the video by performing a set of analyses on each video frame, each analysis selected to detect a meaningful content in the video; arranging the candidate key-frames into a set of clusters; selecting one of the candidate key-frames from each cluster in response to a relative importance of each candidate key-frame.

Because the application of key frame extraction can vary significantly, for example, in terms of the desired number of key frames, it is often desirable to implement a flexible framework capable of producing a scalable video representation. The optimal number of relevant key frames is highly dependent on the video complexity. Complexity is a function of many features: camera motion, scene content, action and interaction between moving objects, image quality (IQ) due to lightning and camera setting, and so on. The video duration is also a parameter that could drive the video complexity: a longer movie clip is likely to contain more events and therefore demands more key frames.

One also need to define the best criteria of representativeness, and then determine what features can be used to obtain the ‘best’ key frames given the input data. Different features, such as those used in US 2005/0228849, vary significantly in terms of their effectiveness and computational cost. It is desirable to use as fewer features as possible to achieve a reasonable performance with reasonable speed.

Furthermore, because video clips taken by consumers are unstructured, one should rely only on cues related to the cameraman's general intents, i.e., camera and object motion descriptors. Rules applicable only to specific content only have limited use and need advance information about the video content.

Consequently, it would be desirable to design a system that is reliable, efficient, regardless of the image content.

SUMMARY OF THE INVENTION

The present invention is directed to overcoming one or more of the problems set forth above. A method according to the present invention analyzes a digital video clip captured by a camera to determine candidate frames for subsequent key frame selection, by:

a. providing a camera motion sensor in the camera so that information is provided during image capture regarding camera motion including translation of the scene or camera, or scaling of the scene;

b. forming a plurality of video segments based on the global motion estimate and labeling each segment in accordance with a predetermined series of camera motion classes; and

c. extracting key frame candidates from the labeled segments and computing a confidence score for each candidate by using rules corresponding to each camera motion class and a rule corresponding to object motion.

One aspect of the present invention focuses on motion-based descriptors as the most effective and efficient features. Camera and object motions are estimated and used to derive a set of motion descriptors. The video clip is divided into a set of homogeneous segments based on the major types of camera motion (pan, zoom . . . ). Dedicated rules allow candidate key frames to be extracted from each segment. The use of complementary descriptors based on image quality (IQ) or semantic analysis (e.g., skin, face or expression) would enable more satisfactory results at an additional expense.

An important feature of the invention is ranking candidates. A confidence score is attached to each candidate and is later used to rank all candidates in an order of relevance. This leads to a scalable framework such that we are able to obtain an arbitrary subset of key frames from candidates given the need of a particular application.

These and other aspects, objects, features and advantages of the present invention will be more clearly understood and appreciated from a review of the following detailed description of the preferred embodiments and appended claims, and by reference to the accompanying drawings.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 is a block diagram illustrating an overview of the key frame extraction method according to the present invention;

FIG. 2 shows an illustration of a video clip containing several camera motion classes and object motion classes, along with desired key frame extraction in response to such motion, in accordance with the interpolation detection method shown in FIG. 1;

FIG. 3 shows a summary of the rules for key frame extraction in response to the camera motion classification of the present invention;

FIG. 4 shows an illustration of a video clip for candidate extraction from a pan segment;

FIG. 5 shows an illustration of a video clip for candidate extraction from a pan segment containing pauses in camera motion; and

FIG. 6 shows an illustration of a video clip for candidate extraction from a zoom-in segment; and

DETAILED DESCRIPTION OF THE INVENTION

Because many basic image and video processing algorithms and methods are well known, the present description will be directed in particular to algorithm and method steps forming part of, or cooperating more directly with, the method in accordance with the present invention. Other parts of such algorithms and methods, and hardware or software for producing and otherwise processing the video signals, not specifically shown, suggested or described herein can be selected from such materials, components and elements known in the art. In the following description, the present invention will be described as a method typically implemented as a software program. Those skilled in the art will readily recognize that the equivalent of such software can also be constructed in hardware. Given the system as described according to the invention in the following materials, software not specifically shown, suggested or described herein that is useful for implementation of the invention is conventional and within the ordinary skill in such arts.

It is instructive to note that the present invention utilizes a digital video which is typically either a temporal sequence of frames, each of which is a two-dimensional array of red, green, and blue pixel values or an array of monochromatic values corresponding to light intensities. However, pixel values can be stored in component forms other than red, green, blue, can be compressed or uncompressed, and can also include other sensory data such as infrared. As used herein, the term digital image or frame refers to the whole two-dimensional array, or any portion thereof that is to be processed. In addition, the preferred embodiment is described with reference to a typical video of 30 frames per second, and a typical frame resolution of 480 rows and 680 columns of pixels, although those skilled in the art will recognize that digital videos of different frame rates and resolutions can be used with equal, or at least acceptable, success. With regard to matters of nomenclature, the value of a pixel of a frame located at coordinates (x,y), referring to the xth row and the yth column of the digital image, shall herein comprise a triad of values [r(x,y), g(x,y), b(x,y)] respectively referring to the values of the red, green and blue digital image channels at location (x,y). In addition, a frame is identified with a time instance t.

Referring to FIG. 1, there is shown an overview block diagram of the present invention. An input video clip 10 first undergoes global motion estimation 20. Based on the estimated global motion, the video clip 10 is then divided through video segmentation 30 into a plurality of segments, each segment 31 corresponding to one of a pre-determined set of camera motion classes 32, including pan (left or right), zoom-in, zoom-out, fast pan, and fixed (steady). For each segment 31, key frame candidate extraction 40 is performed according to a set of pre-determined rules 41 to generate a plurality of candidate key frames 42. For each candidate frame, a confidence score is also computed to rank all the candidates 42 in an order of relevance. Final key frame selection 50 is performed according to a user-specified total number 51 and the rank ordering of the candidates. In a preferred embodiment of the present invention, the final key frames 52 include at least the highest ranked frame in each segment 31.

Because video clips taken by consumers are unstructured, rules applicable only to specific content only have limited use and also need advance information about the video content for them to be useful. In general, one can only rely on cues related to the cameraman's general intents. Camera motion, which usually corresponds to the dominant global motion, allows a prediction of the cameraman's intent. A “zoom in” indicates that he has an interest in a specific area or object. A camera “pan” indicates tracking a moving object or scanning an environment. Finally, a rapid pan can be interpreted as a lack of interest or a quick transition toward a new region of interest (ROI). The secondary or local motion is often an indication of object movements. These two levels of motion description combine to provide a powerful way for video analysis.

In a preferred embodiment of the present invention, a digital camera records, during capture, camera motion information with the help of an integrated camera motion sensor. An example of such a sensor is an accelerometer, which can sense camera orientation and camera translation such that camera pan information can be recorded along with the related video frames as camera motion metadata by using, for instance, the multimedia content description interface in MPEG-7. Furthermore, apparent camera zoom caused by the cameraman's own movement can be recorded this way. In addition, camera zoom operations that are powered by a motor can easily be recorded along with the related video frames as camera motion metadata. According to the present invention, such recorded information is later extracted in a reading camera motion metadata step 20 as a proxy for the camera motion.

Let θ denote the motion-based description vector. Its first 3 components correspond to the camera motion and can be derived either from camera metadata or global motion estimation. This descriptor relies on the translation parameters a1 and a2, and the global divergence (scaling) div. The last descriptor evaluates the amount and the distribution of secondary motion. We refer to secondary motion as the remaining displacement not accounted for by the global motion model. Such spatio-temporal changes are mainly due to objects moving within the 3D scene. The Displaced Frame Difference (DFD) corresponds to the residual motion once the camera motion is compensated. We also combine spatial information (the average distance of the secondary motion to the image center) and the area percentage of the secondary motion. The fourth component of θ is given by:

obj = ω dtc · 1 N Λ p Λ th Hyst [ DFD ( p ) ] ( 1 )

The function thHyst relies on a hysteresis threshold, NΛ is the number of active pixels p, and the normalized linear function Wdtc favors centrally located moving areas.

A video can be characterized in terms of camera motion and object motion. Camera motion is fairly continuous and provides a meaningful partition of a video clip into homogeneous segments in step 30 of FIG. 1. Object activity is an unstable but still useful feature. Referring to FIG. 2, this example video clip consists of the following sequence of camera motion: pan (environment), zoom-in, zoom-out, fast pan, fixed, pan (tracking object), and fixed. Note that a “zoom in” can be caused by a mechanical/optical action from the camera, by the motion of the cameraman (towards the object), or by the movement of the object (towards the camera). However, they are equivalent from an algorithm prospective as “apparent” zoom-in.

As for object motion, the example video clip in FIG. 2 consists of the following sequence of object motion: no object motion, high object motion, and finally low object motion. Note that the boundaries of the object motion segments do not necessarily coincide with the boundaries of the camera motion.

Continuing the reference to FIG. 2, according to the present invention, rules are formulated and confidence functions are defined to select candidate frames for each segment in step 40 of FIG. 1. For the first segment, which is a pan, it would be desirable to select two key frames to span the environment (as marked). For the subsequent zoom-in and zoom-out segments, a key frame should be selected at the end of each segment when the zooming action stops. It is usually not necessary to extract a key frame for the fast pan segment because it is merely transition without any attention paid. Although object motion starts during the latter stage of the fast pan, it is only necessary to extract key frames once the camera becomes steady. One key frame should be extracted as the camera pans to follow the moving object. Finally, as the object moves away from the steady camera, another key frame is selected.

The rules used in the above example are of a general purpose in nature. They do not rely on any semantic information on what the object is, what the environment is, or what the object motion is. Therefore, they can be applied to any other video clips. These generic rules are summarized in FIG. 3.

The present invention distinguishes four camera motion-based classes: “pan,” “zoom in,” “zoom out,” and “fixed.” Note that “tilt” is handled in the same way as “pan” and is treated as the same class (without straightforward modification). Also note that the descriptor obj is not used during video segmentation, which involves applying adaptive thresholds to the scaling and translation curves over time. In the following, detailed descriptions are provided for each camera motion class.

A slow camera pan takes more time to scan a significant area. It seems appropriate to make the segmentation threshold depend on the pan segment's length l, but it is a chicken-and-egg problem because one needs to segment the translation data first to know the length itself. To overcome this problem, a small translation threshold value is used to provide a rough segmentation. There would be no need to extract a pan segment if the camera view does not change significantly. The adaptive threshold thpan is lower when dealing with longer pan. In a preferred embodiment of the present invention, thpan is defined as the unit amount of camera translation required to scan a distance equal to the frame width w multiplied by a normalized coefficient γ that represents a value beyond which the image content is considered to be different enough.

There exists a strong redundancy over time. To save computing time, it is advantageous not to estimate motion for every frame. Instead, a constant temporal sampling rate is maintained over time regardless of the capture frame rate. Let ts denote the temporal subsampling step (the capture frame rate divided by a fixed number of frame samples per second). The time reference attached to the video is denoted as

0 represents the physical time. The second time reference, denoted 1, is related to the subsampled time. Thus,
l′·t s·thpan =γ·w  (2)

The number of frames N is equal to l′.ts, where the duration l′ is considered in

1. Finally, the adaptive threshold is

th pan = γ · w l · t s ( 3 )

A similar method is used to segment the scaling curve. In this case, there is no need to consider a minimal distance to cover but instead a minimum zoom factor. If the scaling process is short, its amplitude must be high enough to be considered. In reference

1, the scaling factor is generalized to

f zoom = t l [ 1 + div ( t ) ] t s ( 4 )

If div(t) is assumed to be the threshold thzoom and constant over time, this expression can be compared to a desired total scaling factor Ys, reflecting the entire zoom motion along a given segment of length l′:
([1+thzoom]t s )l′=γs  (5)

Therefore, the adaptive zoom threshold is given by

th zoom = exp ( ln γ s l t s ) - 1 ( 6 )

The KF candidates form a fairly large set of extracted frames, each of which is characterized by a confidence value. Although such a value differs from camera motion class to class, it is always a function of the descriptor's robustness, the segment's length, the motion descriptor's magnitude, and the assumptions on the cameraman's intent.

In the present invention, high-level strategies are used to select candidates. They are primarily based on domain knowledge. A zoom-in camera operation generally focuses on a ROI. It can be caused by a mechanical/optical action from the camera, movement of the cameraman, or movement of the object. These scenarios are equivalent from the algorithm's perspective as apparent zoom-in. It is desirable to focus on the end of the motion when the object is closest.

Typically, a camera pan is used to capture the environment. Tracking moving objects can also cause camera translations similar to a pan. One way to differentiate between the two scenarios is to make use of the object motion descriptor obj. However, its reliability depends on the ability to compensate for the camera motion. KF candidates are extracted based on the local motion descriptor and the global translation parameters. Camera motion-dependent candidates are obtained according to a confidence function dependent on local translation at a minimal and cumulative panning distance. Other candidates are frames with large object motion.

Finally, for a “fixed” or steady segment, in one embodiment of the present invention, it is reasonable to simply choose the frame located at the midpoint of the segment. Preferred embodiments should use information from additional cues, including image quality (e.g., sharpness, contrast) or semantic descriptors (e.g. facial expression) to select the appropriate frame.

In a preferred embodiment of the present invention, the main goal is to span the captured environment by a minimum number of KF. Because scene content in a consumer video is rarely static, one also needs to consider large object motion. Covering the spatial extent and capturing object motion activity are quite different in nature, and it is nontrivial to choose a trade-off between them. Certainly, a lack of object motion signifies that the cameraman's intent was to scan the environment. In addition, a higher confidence score is assigned to candidates based on the cumulative distance.

To reduce spatial overlap, a probability function dspat is formulated as a function of the cumulative camera displacements. It is null at the segment's onset and increases as a function of the cumulative displacements. The scene content is judged different enough when dspat reaches 1. Once dspat reaches 1, its value is reset to 0 before a new process starts again to compute the cumulative camera displacements. To avoid a sharp transition, its value decreases rapidly according to a Gaussian law to 0 (for instance within the next 3 frames). Note that the cumulative camera displacement is approximated because the camera motion is computed only every ts frames. FIG. 4 shows top candidate frames extracted by using only dspat. Each frame contains distinct content, i.e., to miss any one of them would be to miss part of the whole landscape.

It is worthwhile considering the cameraman's subtler actions. It is noticed that a pause or slow-down in pan often indicates a particular interest, as shown in FIG. 4. It makes sense to assign higher importance to such areas that are local translation minima using the probability function dknow=G(μ, σ), where the function G is a Gaussian function, with μ as the location of local minimum and σ the standard deviation computed from the translation curve obtained upon global motion estimation. Example candidate frames extracted from function dknow are shown in FIG. 4. Because the candidate frames obtained from dspat and dknow can be redundant, one needs to combine dspat and dknow using a global confidence function dpan:
d pan1 d spat2 d know  (7)
with α12=1, such that dpan lies between 0 and 1. Typically, one does not favor either criterion by selecting α12=0.5.

Referring to FIG. 4, candidates are extracted from a pan segment where the pan speed is not constant (as indicated by the ups and downs in the camera translation curve in the middle row). In the top row, six frames are extracted to span the environment while reducing their spatial overlap. In the bottom row, additional five frames are selected according to the minimum points in the translation curve.

Referring now to FIG. 5, there is shown an example of the function dpan, with candidates extracted from a pan segment. Confidence values dpan are used to rank candidate frames. Modes between 0 and 0.5 only display a high percentage of new content, while modes with values greater than 0.5 correspond to a high percentage of new content and are also close to a translation minimum (pan pause). Function dpan enables us to rank such candidate frames.

Fast pan represents either a transition toward a ROI or the tracking of an object in fast motion. In both cases, frames contain severe motion blur and therefore are not useful. It makes sense not to extract KF from such segments. A normalized confident coefficient c based on the translation values is introduced. In a preferred embodiment of the present invention, the coefficient c is reshaped by a sigmoid function:

c ( ω ) = 1 1 + 4 k ( ω - th High ) ( 8 )
where k is the slope at the translation threshold thHigh, and c(thHigh)=0.5. The coefficient c acts as a weighting factor for dpan:
d pan =C(ω)└α1 d spat2 d know┘  (9)

The coefficient c is close to 1 for small translation, decreases around thHigh according to the parameter k, and eventually approaches 0 for large translations.

Candidate selection from a zoom segment is driven by domain knowledge, i.e., KF should be at the end of a zoom segment. The confidence function dzoom can be affected by translation because large pan motion often causes false scaling factor estimates. Similarly to Eq. 8, let cpan denote a sigmoid function that features an exponential term based on the difference between the Euclidian norm of the translation component ω0(t), t being the time associated with the maximal zoom lying within the same segment of the candidate key frame, and a translation parameter trMax (which can be different from thHigh).

The coefficient cpan provides a measure of the decrease in the confidence of the scaling factor when large pan occurs. A high zoom between two consecutive frames is unlikely due to the physical limits of the camera motor. Even though an object might move quickly toward the camera, this would result in motion blur. In a preferred embodiment of the present invention, the maximal permitted scaling factor ths, between two adjacent frames is set to 0.1 (10%), and the ƒzoom factor introduced in Eq. 4 is modified to:

f zoom = t l Ξ ( 1 + div ( t ) , th s ) [ 1 + div ( t ) ] t s - 1 ( 10 )
where the step function is

Ξ ( x , a ) = { 0 if x a x if x < a .

Finally, after applying normalization function N, Eq. 10 can be rewritten as

f zoom = t l k N ( Ξ [ 1 + div ( t ) ] [ 1 + div ( t ) ] t s ) ( 11 )
and the confidence function dzoom for a zoom candidate is
d zoom =c pan·ƒzoom  (12)

Referring now to FIG. 6, there is shown an example of candidate extraction from a series of zoom-in segments. The top row is the plot for (apparent) camera scaling. The bottom row displays the candidate frames rank ordered according to the confidence function dzoom. The actual locations of these candidates are marked in the scaling curve.

Zoom-out segment is processed in a similar fashion, where candidates are extracted at the end of the segment. However, even though a zoom-out operation could be of interest because it captures a wider view of the environment, extracting a candidate key frame from a zoom-out segment is often redundant. The subsequent segment generally contains frames with similar content. In the present invention, a single candidate frame is extracted at the end of a zoom-out segment, but it will be compared to the key frame(s) extract in the next segment to remove any redundancy. To confirm any redundancy, the simplest metrics are histogram difference and frame difference. In a preferred embodiment of the present invention, each frame is partitioned into the same number L of blocks of size M×N, color moments (mean and standard deviation) are computed for each block. The corresponding blocks are compared in terms of their color moments. Two blocks are deemed similar if the distance between the color moments is below a pre-determined threshold. Two frames are deemed similar if the majority (e.g., 90%) of the blocks are similar.

Candidates are also selected based on object motion activity, which can be inferred from the remaining displacement (secondary motion) that is not accounted for by the global motion model. Such spatio-temporal changes are mainly due to objects moving within the 3D scene. Large object motion is often interesting. Therefore, local maximum values of the descriptor obj provide a second set of candidates. Note that their reliability is often lower, compared to camera motion-driven candidates. For example, high “action” values can occur when motion estimation fails and do not necessarily represent true object motion.

There are at least two ways of quantifying secondary motion. One can use the final data values after the M-estimator to compute the deviation from the estimated global motion model, as taught by J.-M. Odobez and P. Bouthemy. Another way is to compensate each pair of frames for the camera motion. Motion compensation is a way of describing the difference between consecutive frames in terms of where each section of the former frame has moved to. The frame l at time t+dt is compensated for the camera motion and object motion is given by Eq. 1.

The confidence function for object motion in a “fixed” segment is a function of its length. A long period without camera motion indicates particular interest of the cameraman. First, the segment length lfix (in reference

1) is rescaled as a percentage of the total video duration such that lfix ε[0,100]. Moreover, it seems reasonable to assume that the gain in interest should be higher from a 1-second to a 2-second segment, than between a 10-second and a 12-second segment. In other words, the confidence function dfix (obj) increases in a non-linear fashion. In a preferred embodiment of the present invention, this observation is modelled by x/(1+x). Therefore,

d fix ( obj ) = l fix · obj 1 + ( l fix · obj ) . ( 13 )

The confidence value for object motion in a “pan” segment is generally lower because the object motion is in the presence of large camera motion. The confidence score is related to the translation amount during the pan: higher confidence is generally associated to object motion-based candidates during small translation. In a preferred embodiment of the present invention, a similar function is used with modification:

d pan ( obj ) = 10 a i th pan · obj 1 + ( 10 a i th pan · obj ) ( 14 )
where the index i of the translation parameter a is either 1 or 2 (for the horizontal and vertical axes).

The confidence value for object motion in a “zoom” segment is set to zero because object motion within a zoom segment is highly unreliable. Therefore, dzoom (Obj)=0 and no candidate is extracted based on object motion.

Although the present invention is embodied primarily using camera motion and object motion cues, those skilled in the art can use complementary descriptors, such as image quality (IQ), semantic analysis (e.g., skin, face, or expression) to improve the results at additional expense, without deviating from the scope of the present invention.

In the last step 50 of FIG. 1, final key frames 52 are selected from the initial candidates 42. The confidence value of each candidate enables rank ordering. To space out KF, at least one key fame (the highest ranked candidate) is extracted per segment unless its confidence value is too low. To fill in the user-specified number of key frames NKF, the remaining candidates with the highest confidence values are used. If two candidates are too close in value, only the one with the higher confidence value is retained. Preferred embodiments should use information from additional cues, including image quality (e.g., sharpness, contrast) or semantic descriptors (e.g. facial expression) to select the appropriate frame.

The present invention has been described with reference to a preferred embodiment. Changes can be made to the preferred embodiment without deviating from the scope of the present invention. Such modifications to the preferred embodiment do not significantly deviate from the scope of the present invention.

PARTS LIST

  • 10 input digital video
  • 20 global motion estimation
  • 30 video segmentation
  • 31 video segments
  • 32 camera motion classes
  • 40 candidate frame extraction
  • 41 rules
  • 42 candidate frames
  • 50 key frame selection
  • 51 key frame number
  • 52 key frames
Patent Citations
Cited PatentFiling datePublication dateApplicantTitle
US5793886 *Dec 19, 1996Aug 11, 1998Eastman Kodak CompanyMethod of adjusting image pixel values within an image sequence by interpolation of target cumulative histograms for images to be adjusted
US6122320 *Mar 12, 1998Sep 19, 2000Cselt-Centro Studi E Laboratori Telecomunicazioni S.P.A.Circuit for motion estimation in digitized video sequence encoders
US6549643 *Nov 30, 1999Apr 15, 2003Siemens Corporate Research, Inc.System and method for selecting key-frames of video data
US6584155 *Dec 26, 2000Jun 24, 2003Kabushiki Kaisha ToshibaMethod and system for estimating motion vector
US6721454 *Jul 1, 1999Apr 13, 2004Sharp Laboratories Of America, Inc.Method for automatic extraction of semantically significant events from video
US7432940 *Oct 11, 2002Oct 7, 2008Canon Kabushiki KaishaInteractive animation of sprites in a video production
US7561177 *Jul 19, 2004Jul 14, 2009Hewlett-Packard Development Company, L.P.Editing multiple camera outputs
US7612832 *Mar 29, 2005Nov 3, 2009Microsoft CorporationMethod and system for video clip compression
US20020018594 *Apr 20, 2001Feb 14, 2002Mitsubishi Electric Research Laboratories, Inc.Method and system for high-level structure analysis and event detection in domain specific videos
US20040088723 *Nov 1, 2002May 6, 2004Yu-Fei MaSystems and methods for generating a video summary
US20040136460 *Jan 15, 2003Jul 15, 2004Zhang HongjiangMethod and system for extracting key frames from video using a triangle model of motion based on perceived motion energy
US20050206751 *Mar 19, 2004Sep 22, 2005East Kodak CompanyDigital video system for assembling video sequences
US20050228849 *Mar 24, 2004Oct 13, 2005Tong ZhangIntelligent key-frame extraction from a video
US20060093040 *Dec 17, 2005May 4, 2006Microsoft CorporationMethod and system for extracting key frames from video using a triangle model of motion based on perceived motion energy
US20060177207 *Feb 3, 2006Aug 10, 2006Samsung Electronics Co., Ltd.Method and apparatus for expressing amount of camera motion
US20060185431 *Feb 23, 2005Aug 24, 2006Henryk BireckiCamera motion detection system
US20060193387 *Jan 26, 2006Aug 31, 2006Canon Kabushiki KaishaExtracting key frames from a video sequence
US20060257048 *May 12, 2005Nov 16, 2006Xiaofan LinSystem and method for producing a page using frames of a video stream
US20070182861 *Feb 3, 2006Aug 9, 2007Jiebo LuoAnalyzing camera captured video for key frames
US20070237225 *Mar 30, 2006Oct 11, 2007Eastman Kodak CompanyMethod for enabling preview of video files
Non-Patent Citations
Reference
1"A Novel Video Key Frame Extraction Algorithm" by T. M. Liu, et al., IEEE Trans. Cir. Sys. Video Techno, 13(10), 2003, pp. 1006-1013.
2"A Stochastic Framework for Optimal Key Frame Extraction from MPEG Video Databases" by Yannis S. Avrithis et al., Computer Vision and Image Understanding, 75(½), 1999, pp. 3-24.
3"Automatically Extracting Highlights for TV Baseball Programs" by Yong Rui et al., ACM Multimedia 2000, pp. 105-115.
4"Event Detection and Summarization in Sports Video" by Baoxin Li et al., IEEE Workshop on Content-based Access of Image and Video Libraries (CBAIVL), 2001, pp. 132-140.
5"Robust Multiresolution Estimation of Parametric Motion Models Applied to Complex Scenes" by Jean-Marc Odobez et al., IRISA Campus Universitaire de Beaulieu, ISSN 1166-8687.
6 *Wolf, Key Frame Selection by Motion Analysis, 1996, IEEE International Conference on Acoustics, Speech, and Signal Processing, vol. 2, pp. 1228-1231.
Referenced by
Citing PatentFiling datePublication dateApplicantTitle
US8432965 *May 25, 2010Apr 30, 2013Intellectual Ventures Fund 83 LlcEfficient method for assembling key video snippets to form a video summary
US8446490 *May 25, 2010May 21, 2013Intellectual Ventures Fund 83 LlcVideo capture system producing a video summary
US8670593 *May 25, 2011Mar 11, 2014Samsung Electronics Co., Ltd.Apparatus and method for dynamically controlling integration time of depth camera for accuracy improvement
US20110292245 *May 25, 2010Dec 1, 2011Deever Aaron TVideo capture system producing a video summary
US20110293018 *May 25, 2010Dec 1, 2011Deever Aaron TVideo summary method and system
US20120106803 *May 25, 2011May 3, 2012Samsung Electronics Co., Ltd.Apparatus and method for dynamically controlling integration time of depth camera for accuracy improvement
Classifications
U.S. Classification375/240.16, 348/700
International ClassificationH04N3/36, H04N5/14
Cooperative ClassificationH04N19/00921, H04N19/00593, H04N5/145, G11B27/28, G11B27/3027
European ClassificationH04N7/26M2G, G11B27/28, H04N5/14M2, G11B27/30C, H04N7/26P8
Legal Events
DateCodeEventDescription
Mar 8, 2013ASAssignment
Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNOR:EASTMAN KODAK COMPANY;REEL/FRAME:029939/0508
Effective date: 20130211
Owner name: APPLE INC., CALIFORNIA
Feb 1, 2013ASAssignment
Owner name: FPC INC., CALIFORNIA
Owner name: NPEC INC., NEW YORK
Free format text: PATENT RELEASE;ASSIGNORS:CITICORP NORTH AMERICA, INC.;WILMINGTON TRUST, NATIONAL ASSOCIATION;REEL/FRAME:029913/0001
Owner name: KODAK AMERICAS, LTD., NEW YORK
Effective date: 20130201
Owner name: KODAK PORTUGUESA LIMITED, NEW YORK
Owner name: KODAK REALTY, INC., NEW YORK
Owner name: QUALEX INC., NORTH CAROLINA
Owner name: EASTMAN KODAK COMPANY, NEW YORK
Owner name: FAR EAST DEVELOPMENT LTD., NEW YORK
Owner name: PAKON, INC., INDIANA
Owner name: EASTMAN KODAK INTERNATIONAL CAPITAL COMPANY, INC.,
Owner name: CREO MANUFACTURING AMERICA LLC, WYOMING
Owner name: KODAK (NEAR EAST), INC., NEW YORK
Owner name: LASER-PACIFIC MEDIA CORPORATION, NEW YORK
Owner name: KODAK AVIATION LEASING LLC, NEW YORK
Owner name: KODAK IMAGING NETWORK, INC., CALIFORNIA
Owner name: KODAK PHILIPPINES, LTD., NEW YORK
Feb 21, 2012ASAssignment
Effective date: 20120215
Free format text: SECURITY INTEREST;ASSIGNORS:EASTMAN KODAK COMPANY;PAKON, INC.;REEL/FRAME:028201/0420
Owner name: CITICORP NORTH AMERICA, INC., AS AGENT, NEW YORK
May 3, 2006ASAssignment
Owner name: EASTMAN KODAK COMPANY, NEW YORK
Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNORS:JUO, JIEBO;PAPIN, CHRISTOPHE E.;SCHILDKRAUT, DENIZ E.;AND OTHERS;REEL/FRAME:017842/0255;SIGNING DATES FROM 20060407 TO 20060426
Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNORS:JUO, JIEBO;PAPIN, CHRISTOPHE E.;SCHILDKRAUT, DENIZ E.;AND OTHERS;SIGNING DATES FROM 20060407 TO 20060426;REEL/FRAME:017842/0255